not important/ testing purpuses

Added some spaces in the text

Author: Yulia Bezsudnova

docs/tutorials/opm/evoked/index.md

@@ -98,10 +98,10 @@ mD = spm_opm_amm(S);
 ``` 
 ### Spatio-temporal models of interference and brain signal 
 
-If the source of magnetic interference is quite near the array then the low order spatial models will not be able to remove the interference. This problem is further exacerbated by the presence of sensor non-linearities such as calibration and orientation errors. This kind of interference can be identified using a temporal subspace intersection (implemented using CCA). This step requires setting the argument `S.corrLim` which is the the threshold(value between 0 and 1) at which one considers two subspaces to have intersected. 
+If the source of magnetic interference is quite near the array then the low order spatial models will not be able to remove the interference. This problem is further exacerbated by the presence of sensor non-linearities such as calibration and orientation errors. This kind of interference can be identified using a temporal subspace intersection (implemented using CCA). This step requires setting the argument `S.corrLim` which is the the threshold (value between 0 and 1) at which one considers two subspaces to have intersected. 
 
 ??? info "How does subspace intersection work?" 
-	`spm_opm_amm` defines 3 subspaces. A brain space using internal spheroidal harmonics (of default order `S.li=9`), an interference space using external spheroidal harmonics (of default order `S.li=2`) and an intermediate space that contains the data that is not well modelled by the internal or external harmonics. If the origin of magnetic interference is spatially close by it will not be well modelled by the external harmonics and will be present in the intermediate space. However, due to partial spatial correlation between complex interference and brain signal part of this interference will also be present in the brain signal space. If we assume that over long time scales (5-10s) brain signals(as observed by OPMs) are poorly temporally correlated any high temporal correlations (temporal subspace intersection) between the brain space and the intermediate space are most likely a reflection of magnetic interference. This interference is then readily removed using linear regression. 
+	`spm_opm_amm` defines 3 subspaces. A brain space using internal spheroidal harmonics (of default order `S.li=9`), an interference space using external spheroidal harmonics (of default order `S.li=2`) and an intermediate space that contains the data that is not well modelled by the internal or external harmonics. If the origin of magnetic interference is spatially close by it will not be well modelled by the external harmonics and will be present in the intermediate space. However, due to partial spatial correlation between complex interference and brain signal part of this interference will also be present in the brain signal space. If we assume that over long time scales (5-10s) brain signals (as observed by OPMs) are poorly temporally correlated any high temporal correlations (temporal subspace intersection) between the brain space and the intermediate space are most likely a reflection of magnetic interference. This interference is then readily removed using linear regression. 
 
 
 ```matlab
@@ -111,7 +111,7 @@ S.corrLim = .98;
 mD = spm_opm_amm(S);
 ``` 
 ??? info "How do I pick a value for S.corrLim?"
-	Setting a lower value for `S.corrLim` will result in more aggressive cleaning of the data but will increase the risk of removing interesting brain signal. The factors that influence this decision are the order of the internal harmonics (`S.li`), the SNR of the brain response, the time scale at which physiological correlations exist, the length of the time window analysed and the number of channels in your array. As such it is not possible to extrapolate any heuristics one might have developed from using other temporal subspace intersection methods such as TSSS. However, the code has been developed so that  values greater than `0.98` are generally safe for arrays of less than 200 channels. 
+	Setting a lower value for `S.corrLim` will result in more aggressive cleaning of the data but will increase the risk of removing interesting brain signal. The factors that influence this decision are the order of the internal harmonics (`S.li`), the SNR of the brain response, the time scale at which physiological correlations exist, the length of the time window analysed and the number of channels in your array. As such it is not possible to extrapolate any heuristics one might have developed from using other temporal subspace intersection methods such as tSSS. However, the code has been developed so that  values greater than `0.98` are generally safe for arrays of less than 200 channels. 
 
 
 !!! note "How do I choose which model to use?"